skills/duthaho/skillhub/verdict/Gen Agent Trust Hub

verdict

Pass

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to use a shell command (date +%F) to retrieve the current system date for output file naming. This is a benign use of command execution for environment discovery.
  • [EXTERNAL_DOWNLOADS]: The skill fetches data from several external services, including the GitHub REST API and Hacker News (via the Algolia API). While these are well-known services, the retrieval of external content for processing constitutes an external data dependency.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted content from the public web (e.g., repository READMEs, forum discussions) to generate its recommendations.
  • Ingestion points: External research data is gathered by sub-agents as described in SKILL.md (Step 1).
  • Boundary markers: The skill lacks explicit delimiters or instructions to treat ingested research data as untrusted, increasing the risk that the agent may follow instructions embedded within that data.
  • Capability inventory: The skill has the ability to execute shell commands and write files to the local directory (out/verdicts/).
  • Sanitization: No specific sanitization or filtering logic is provided to clean the external content before it is processed by the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 25, 2026, 04:21 PM
Security Audit — agent-trust-hub — verdict